Sign up Calendar Latest Topics Donate
 
 
 


Reply
  Author   Comment  
Vibe-Rater

Avatar / Picture

Sr. Member / Moderator / Supporter
Registered:
Posts: 1,392
Reply with quote  #1 
Hi Rusty,

I've been getting some qoutes for a triax. CTC and Emerson / CSI.  They are really expensive.

So I managed to borrow one.

Wondering if you can give me a triax for dummy's explanation of how to set up in the database and how to cope with the different directions. The suitable mounting locations will not always have the flats in the same orientations. Does this mean that for every machine you have to change setup to suit the orientation of the triax.. probably yeah.

I recently monitored a testbed epicyclic gearbox and having to use single axis worked well but time would be saved on the 15 rpm points. rgds
fburgos

Sr. Member
Registered:
Posts: 659
Reply with quote  #2 
I'm also interested in this, I'm stating my db and plannig to be full triax hope its not a mistake because triax is expensive and i only have one.

The CSI its a two pole "for curved surfaces" I'm trying to use A horizontal and peakvue, B Axial, C Vertical.

Personally I hate video tutorial, but i use this one of DBUtly to change "all" the database group and signal channel.



I try to keep things organized same way, then mayority of locations share the same configuration I would note down the "special" configuration.

first ill do the DButly thing to all the area, then and change those "special" on DBase.


If you ask me I think CTC best because you can use any mounting method you like.
RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,798
Reply with quote  #3 
The CSI triax has a built-in 2-rail magnet which is surprisingly versatile, and being integral to the sensor body, eliminates an "interface" detrimental to PeakVue readings.

Below is the nomenclature CSI uses: 

CSI Triax.png 
 
I am not sure how other manufacturer designate the axes - I don't even know "the right hand rule" - but it really doesn't matter.  All you need to think about is how A, B, & C are oriented for the point you are measuring, and how your cable is wired.  The cable I purchased from CTC was set up so that the wires (A-B-C, 1-2-3, or X-Y-Z ??) match the sensor output.  Of course this is not actually a "requirement" but it simplifies things.  My cable matches the CSI triax, but I think this was just a "happy accident" because I didn't order what I thought I did.  But you absolutely (IMO) want the cable to match the sensor, regardless of how the sensor is set up.

My approach is to always orient the rails horizontally -- i.e., never vertically.  If the surface is round, it's a no brainer.  If mounting on a flat surface, I just keep the same orientation.

Mounted in a normal horizontal position (side of bearing):  A = horizontal, C = vertical, B = axial
Mounted in a vertical postion (top of bearing for instance): A = vertical, C = horizontal, B = axial
Mounted in an axial position: A = axial, B = horizontal, C = vertical

It's easiest to simply have the accelerometer in hand as you set up the point, and just think about how it will be mounted.  To make mine easier to read, I added labels for B and C, on both sides.  I just always mount the triax as I would if it were a curved surface - 'C' is always 'up/down'.

The secret (to me) is to reorder your route points so that the 1st point (on the route) is your primary or A axis. 

Each triaxial location (physical measurement point) gets a different 'group number' when you define your measurement points in the database.  Four a 4-brg machine, you'd have 4 groups if you collected every physical point as triaxial.  But every measurement point does not have to be defined as triaxial.  If no group is defined (i.e., group = '0') for a particular point, it just records the primary (A) reading and ignores B and C.

Suppose I have a large motor, with a round flange. I use MOH, MOV, and MOA for the outboard end of a motor.  I will usually be most interested in the horizontal position, so I mount my 2-rail triax at that location.  The sensor will record: A = MIH, B = MIA, C = MIV.  In my database, the points will always be ordered as H-V-A, so MIH = Chnl 1;  MIV = Chnl 3;  MIA = Chnl 2  and all will be designated Grp 1.  In my 'route' definition, MIH will be the first point (of those 3).  It really doesn't matter which of the other 2 points is 2nd, and with is 3rd.  I walk up, move to that location in the route, and MIH is the first thing I see. I mount the sensor there, push Enter, and all 3 points are collected. I see the overall values briefly, and the meter auto-advances to the next group.

If I also have a PeakVue point defined for the MIH position, it is also Chnl 1.  When I push the button, all 4 are collected.  If I have "extra" points (high Fmax, high resolution, low Fmax, whatever) defined for that position, I just give them all the same Group number, and the appropriate Channel number.  I could have 4 points defined for each position (H,V,A), and all 12 readings are collected with 1 button push.

On Centac air compressors, there is a solid, round flange at the motor outboard bearing, so with a 2-rail the H and A positions are both good locations (V is usually not).  But I find the best point for motor bearing problem detection is MOA, so that's where I mount the triax.  So MOA = Ch. 1, MOH = Ch. 2, MOV = Ch. 3.  In my route definition, MOA will be the first point for that bearing.  It doesn't matter about the order of the Ch 2 and Ch 3 points, except what you prefer if you want to look at each spectrum/waveform after you've collected it.

Just always make sure that the point designated Ch. 1 is the first point you see for that bearing.  And when you set up your channel designations, make sure you have the sensor in hand with the directions (B & C) clearly labeled.  Sorry so long -- it's much simpler to do than it is to explain.

Triax MOA.jpg  Triax MOH.jpg 


__________________
"The trend is your friend"
RustyCas

Avatar / Picture

Admin
Registered:
Posts: 1,798
Reply with quote  #4 
The problem I have with any triax for route work, is that it doesn't always provide the best data for every reading you want to take.  For instance, on Dodge Sleeve-Oil bearings, at the outboard bearing I collect:  Horizontal = side of bearing, flat surface;  Vertical = on top of the "liner bolt";  Axial = end of bearing (machined flat); and CWO is "outboard cooling water line, vertical" (on the metal nipple which screws into the liner; tells me without fail if the liner is "loose", i.e., the liner bolt needs to be re-torqued).  I don't want to collect any of these points as triaxial, so I don't.  I just leave the Group and Channel as '0' and use the sensor as I would any other - orientation of the rails doesn't matter since B and C are ignored.

On the inboard fan bearing, I really want to collect an Axial reading because that's where the "thrust" collar is located.  But there's no room, so I've never collected it.  But with the triaxial, I just put H & A in a group, and H = Ch. 1 (A) and Axial = Ch. 2 (B), and I get a good axial reading.  I just leave V and CWI (cooling water line) as Grp/Ch '0' and collect them as I always have.

I use the triaxial sensor mostly for fans and air compressors.  Don't really use it on my pump routes because there is seldom a good place to attach a magnet -- most of those locations are best grabbed with a "stinger".  So I will switch cables and use my non-triax setup (swapping out a 2-rail, a flat magnet, and a stinger as needed).

But on my fan routes I have some smaller, shaft-mounted fans with no magnet locations on the motors, and some of the larger motors I can't use a magnet.  But I don't want to be constantly switching cables.  This is where the integrated magnet of the triax comes in handy.  It ships with a "pad" which exactly matches the contour of the magnet.  I assume the idea is to purchase these and epoxy them to your machines so that the sensor orientation is always correct and mounting is consistent.  However, this pad is center drilled for a 1/4" bolt, with a recess perfect for an Allen-head screw.  So I simply attached a short steel stinger to the pad using a 1/4-28 Allen-head screw.  I just stick it on as needed, and then take it off for my triaxial, and 'typical' magnet readings.

P5022569.jpg 
P5022570.jpg


__________________
"The trend is your friend"
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.